|
Theories
Newsgroups: rec.games.go,fj.rec.games.go,japan.games.go
Date: 2001-05-11 to 12 LECTURES
Subgames with goals (like connection) are considered by using mathematics to describe otherwise only heuristical strategic analysis precisely. Abstract proof search and other search methods sincerely reduce calculation complexity in computer Go. [The work is impressive and promising. The speaker will continue this research as a professor.]
So far computer Go fails
also because too much human wisdom requires description. Machine learning
tries to overcome this by letting the programs learn Go theory themselves.
However, still it must be defined _what_ should be learned. [Because
of this difficulty quick breakthroughs are unlikely, unless an application
is very sepcialized.]
Psychological studies try to assess ways of information storage by expert players in comparison to weaker players. Different types of storage (e.g. a long-term memory) exist. Human expert skill does not imply skill in general. Research is done on explanations for this. [The psychological studies seem to be profound while much future research as to correct interpretations is required.]
On overview on theory and practice of the EGF rating system and tuning of some of its parameters is given. The system is said to be reliable and extendable to include players of more countries. [The major obstactles of rating systems will not be overcome easily because of, e.g., partially too little data input.]
Asian thinking and philosophy is reflected in Go as a game and in its strategies. E.g. Asian military planning has aspects similar to Go strategy. Go might be a tool for Westerners to gain better access to Asian thinking. [I have not really understood the lecture yet because I am not familar with the underlying philosophical background.]
Anthropological studies suggest that fundamental aspects of shamanism in Eastern Asia can be compared to mythologic foundations of Go. [I will have to read to paper before really understanding any details.]
Modern science and economis have developed powerful methods of scientific study. Often these lack example study objects that are difficult and complex enough for gaining new insight. However, Go is often very suitable in this respect.
Ancient playing material, archeologic and literary evidence suggests that and how Go with systems of setup stones at the beginning of a game changed and finally resulted in even game Go starting from the empty board. [Different theories exist because evidence is rare. E.g. one theory locates the origin of Go in or near to Tibet.]
Actually, it was a lecture
about a method of teaching Go to children relying on a division of all
stones into nets, fish, walls separating others, and useless walls in
between walls.
While teaching in the East heavily relies on examples, teaching in the West tends to use principles. Therefore teaching by Asian professionals in the West can be difficult. Empirical surveys contribute to solving this problem. [Korean professionals are very serious in promoting Go in the West now and even want to surpass other countries.]
Teaching adult beginners has its own difficulties. E.g. an enthusiastic teacher might easily explain too much or miss the particular needs of his pupils. These problems should be studied and a reasonably simplified teaching style applied. [As always, teaching depends on the particular relationship between teachers and pupils. Simplification will not work in all cases.] A method of teaching Go to children similar to that one promoted by the EGF is explained. Essentially children are supposed to be proud of discovering concepts for themselves. Teaching pretends to use Ing rules but in fact uses superko because of simplicity. [The speaker was not aware that the nasty Ing fill-in counting can be replaced by the easier stone counting, which only fills regions surrounded only by one's own stones, then removes pairs of one black and one white stone, and then leaves the score as its remainder.]
[The speech was in Chinese and I have not had time to read the text yet. So I can only state my first impression that teaching teachers is done excessively and on a methodically high level.]
During the last few years
amateurs have contributed a lot to scientific Go theory by discussing
in the Internet.
A particularly excellent
type of middle game moves relies on accurate reading, achieves an important
goal, and ensures that one's own positions unrelated to the goal remain
uneffected from the fight following play of such a move.
For human endgame calculation different, competing value description systems exist, which are not easily compatible to each other. Their merits are discussed. [This is very useful, applied study, although maybe not quite as formal as combinatorial game theory so far.]
It shall be studied how human players of different strengths solve Go problems. Empirical evidence collected in Korea suggests that strong players concentrate on a by experience well chosen narrow search, which can fail in case of eccentric moves. Computers think differently because their brains rely on other advantages. [Much study remains to be done, e.g. also collection of data outside Korea.]
A quick overview on the Database, its means, and presentations and its flexible add-ons in the future was given. [The database is big and powerful and so the text describing might reveal more interesting information, although it does probably hardly cover every aspect.]
Because of time restrictions, only a few problems of translation could be illustrated but it has become clear that they can origin from a large number of different methodical causes. E.g. the object rice measure is not known the West and thus a direct translation has little meaning.
Tournament systems, their practical application, the development and promotion of Go as a sport among others in Yugoslavia was discussed. [Concerning tournament systems only basic information was given. On the other hand, it has become clear that national support of sports can greatly promote the game.]
This was a pathetic report of the positive effect that Go has had on the speaker's life.
From some philosophical point of view one can draw and study parallels between playing Go and education. [I will have to read to paper before really understanding any details.] SHORTCOMINGS Preparation time for the conference was too short and caused a few minor problems like not perfectly edited proceedings. There were no participants from Latin America or New Zealand, maybe because nobody applied. CONCLUSION The conference and all activities
thereafter (visiting Seoul and Baduk academies) were well organized
and brought different cultures and scientific approaches together. It
was a great experience and we can expect much from future conferences.
Korea is at the top of playing and aspires to get at the top of teaching,
researching, and organizing. Considering all the current activities,
achievement of this is not unlikely. However, in view of the conference's
scope every Go player in the world will profit.
|